Kerry Back
BUSI 721, Fall 2022
JGSB, Rice University
The Capital Asset Pricing Model (CAPM) is a theory from the 1960s that has been taught to MBAs for 50 years.
It is widely used to estimate expected returns to compute discount factors for evaluating corporate investment projects.
It does not work empirically.
The hypothesis of the CAPM is that the market portfolio is mean-variance efficient, with borrowing rate = savings rate.
In the usual application in the U.S., the “market portfolio” is the U.S. stock market.
The rationale for mean-variance efficiency is that if each investor chooses a mean-variance efficient portfolio, then each investor holds the tangency portfolio
⇒ market portfolio is the tangency portfolio
Mean-variance efficiency implies that all alphas are zero (otherwise we could mean-variance improve).
With \(\alpha > 0\), \[r-r_{f}=\beta(r_m-r_f)+\epsilon\]
where \(r_{m}\)=market return.
This implies \(\bar{r} - r{f} + \beta(\bar{r}_{m}-r_{f})\) or \(\bar{r}=r_{f}+\beta(\bar{r}_m-r_{f})\)
\(\beta\) > 1 implies higher risk premium.
\(\beta\) < 1 implies lower risk premium.
The data don’t comply with the theory.
A simple example is industry returns. Average returns are unrelated to betas.